• Journal of the Institute of Convergence Signal Processing
• /
• v.10 no.1
• /
• pp.13-22
• /
• 2009

In this paper, we propose the 3D shape reconstruction method that combine the mosaic method and the active stereo matching using the laser beam. The active stereo matching method detects the position information of the irradiated laser beam on object by analyzing the color and brightness variation of left and right image, and acquires the depth information in epipolar line. The mosaic method extracts feature point of image by using harris comer detection and matches the same keypoint between the sequence of images using the keypoint descriptor index method and infers correlation between the sequence of images. The depth information of the sequence image was calculated by the active stereo matching and the mosaic method. The merged depth information was reconstructed to the 3D shape information by wrapping and blending with image color and texture. The proposed reconstruction method could acquire strong the 3D distance information, and overcome constraint of place and distance etc, by using laser slit beam and stereo camera.

• Korean Journal of Digital Imaging in Medicine
• /
• v.14 no.2
• /
• pp.57-61
• /
• 2012

To compare the accuracy of breath-hold magnetic resonance imaging sequences to establish the most effective superparamagnetic iron oxide-enhanced sequence for detection of hepatic metastases. A total of 100 patients(50men and 50women, mean age: 60years) with liver disease(including malignant and benign liver lesions) were investigated at 3.0T machine (GE, General Electric Medical System, Excite HD) with 8Ch body coil. Pulse sequence for MR imaging decided to the FS-T2-FSE-RT(TR/TE/Thick./Freq./Phase=12857ms/100ms/7mm/512/384), MGRE(TR/TE/Thick./Freq./Phase=100ms/9.7ms/7mm/384/288), in-out of phase echo(TR/$TE_1$, $TE_2$/Thick./Freq./Phase=140ms/2.4, 5.8ms/7mm/352/300), Images obtained before the injection of SPIO. Six sequences were optimized for lesion detection: FS-T2-FSE-RT, multigradient recalled echo data image(MGRE), T2-weighted MGRE with an 9.7msec echo time. Images were reviewed independently by five blinded observers. The accuracy of each sequence was measured by using picture archiving communication system analysis. All results were correlated with findings at multidectator computed tomography examination. Differences between the mean results of the six observers were measured by using paired student t-test analysis. Postcontrast T2-weighted MGRE sequences were the most accurate and were significantly superior to postcontrast FS-T2-FSE-RT, T2-weighted MGRE, in-out of phase MR sequences(p < .05). For all lesions that were malignant or smaller than 1 cm, respectively, contrast to noise ratio of pre and postcontrast sequences were -1and -0.3 for T2-weighted FSE, 0.53 and 4.5 in-out of phase, 7, 7.08, 5.08, 3.32, 1.7, 1.16, 0.79, 0.68 for GRE with 2.9, 7.5, 12.1, 16.6, 21.2, 25.8, 30.4, 35.0 TE values. Breath-hold various TE precontrast sequences offer improvement in sensitivity compared with fixed multigradient recalled echo sequences alone.

• Progress in Medical Physics
• /
• v.30 no.1
• /
• pp.22-31
• /
• 2019

Purpose: To develop a new diffusion-based functional MRI (fMRI) sequence to generate apparent diffusion coefficient (ADC) maps in single excitation and evaluate the contribution of b0 signal on neuronal changes. Materials and Methods: A diffusion-based fMRI sequence was designed with single measurement that can acquire images of three directions at a time, obtaining $b=0s/mm^2$ during the first baseline condition (b0_b), followed by 107 diffusion-weighted imaging (DWI) with $b=600s/mm^2$ during the baseline and visual stimulation conditions, and another $b=0s/mm^2$ during the last activation condition (b0_a). ADC was mapped in three different ways: 1) using b0_b (ADC_b) for all time points, 2) using b0_a (ADC_a) for all time points, and 3) using b0_b and b0_a (ADC_ba) for baseline and stimulation scans, respectively. The fMRI studies were conducted on the brains of 16 young healthy volunteers using visual stimulations in a 3T MRI system. In addition, the blood oxygen level dependent (BOLD) fMRI was also acquired to compare it with diffusion-based fMRI. A sample t-test was used to investigate the voxel-wise average between the subjects. Results: The BOLD data consisted of only activated voxels. However, ADC_ba data was observed in both deactivated and activated voxels. There were no statistically significant activated or deactivated voxels for DWI, ADC_b, and ADC_a. Conclusions: With the new sequence, neuronal activations can be mapped with visual stimulation as compared to the baseline condition in several areas in the brain. We showed that ADC should be mapped using both DWI and b0 images acquired with the same conditions.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.10
• /
• pp.3708-3728
• /
• 2021

Indian herbal plants are used in agriculture and in the food, cosmetics, and pharmaceutical industries. Laboratory-based tests are routinely used to identify and classify similar herb species by analyzing their internal cell structures. In this paper, we have applied computer vision techniques to do the same. The original leaf image was preprocessed using the Chan-Vese active contour segmentation algorithm to efface the background from the image by setting the contraction bias as (v) -1 and smoothing factor (µ) as 0.5, and bringing the initial contour close to the image boundary. Thereafter the segmented grayscale image was fed to a leaky capacitance fired neuron model (LCFN), which differentiates between similar herbs by combining different groups of pixels in the leaf image. The LFCN's decay constant (f), decay constant (g) and threshold (h) parameters were empirically assigned as 0.7, 0.6 and h=18 to generate the 1D feature vector. The LCFN time sequence identified the internal leaf structure at different iterations. Our proposed framework was tested against newly collected herbal species of natural images, geometrically variant images in terms of size, orientation and position. The 1D sequence and shape features of aloe, betel, Indian borage, bittergourd, grape, insulin herb, guava, mango, nilavembu, nithiyakalyani, sweet basil and pomegranate were fed into the 5-fold Bayesian regularization neural network (BRNN), K-nearest neighbors (KNN), support vector machine (SVM), and ensemble classifier to obtain the highest classification accuracy of 91.19%.

• 제어로봇시스템학회:학술대회논문집
• /
• 2005.06a
• /
• pp.383-388
• /
• 2005

Depth recovery in robot vision is an essential problem to infer the three dimensional geometry of scenes from a sequence of the two dimensional images. In the past, many studies have been proposed for the depth estimation such as stereopsis, motion parallax and blurring phenomena. Among cues for depth estimation, depth from lens translation is based on shape from motion by using feature points. This approach is derived from the correspondence of feature points detected in images and performs the depth estimation that uses information on the motion of feature points. The approaches using motion vectors suffer from the occlusion or missing part problem, and the image blur is ignored in the feature point detection. This paper presents a novel approach to the defocus technique based depth from lens translation using sequential SVD factorization. Solving such the problems requires modeling of mutual relationship between the light and optics until reaching the image plane. For this mutuality, we first discuss the optical properties of a camera system, because the image blur varies according to camera parameter settings. The camera system accounts for the camera model integrating a thin lens based camera model to explain the light and optical properties and a perspective projection camera model to explain the depth from lens translation. Then, depth from lens translation is proposed to use the feature points detected in edges of the image blur. The feature points contain the depth information derived from an amount of blur of width. The shape and motion can be estimated from the motion of feature points. This method uses the sequential SVD factorization to represent the orthogonal matrices that are singular value decomposition. Some experiments have been performed with a sequence of real and synthetic images comparing the presented method with the depth from lens translation. Experimental results have demonstrated the validity and shown the applicability of the proposed method to the depth estimation.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.42 no.3 s.303
• /
• pp.41-50
• /
• 2005

A novel image mosaicking technique suitable for 3D urban visualization is proposed. It is not effective to apply 2D image mosaicking techniques for urban visualization when, for example, one is filming a sequence of images from a side-looking video camera along a road in an urban area. The proposed method presents the roadside scene captured by a side-looking video camera as a continuous set of textured planar faces, which are termed 'multiple planes' in this paper. The exterior parameters of each frame are first calculated through automatically selected matching feature points. The matching feature points are also used to estimate a plane approximation of the scene geometry for each frame. These planes are concatenated to create an approximate model on which images are back-projected as textures. Here, we demonstrate algorithm that creates efficient image mosaics in 3D space from a sequence of real images.

• Journal of Korea Multimedia Society
• /
• v.15 no.7
• /
• pp.847-858
• /
• 2012

In this paper, we propose a mobile watermarking based on the distortion analysis of display-capture image in a smart phone. We compose a random sequence by utilizing the property of frequency band in the wavelet domain. Then we calculate the CCS (Coefficients Comparative Sum) using the block wavelet coefficients of selected subbands after the wavelet transformation and the random sequence and repeatedly embed a watermark using an insertion threshold for the watermark robustness. For correcting a distortion caused by the display-capture process, we adopt a frame at the outside of watermarked image, then we can equate a watermark synchronization by detecting the frame. And we can improve frame detection ratio by using an iteratively adaptive threshold. A proposed scheme embedded information of 206 bits into standard digital images and it shows an average about 41.42 dB in PSNR. In watermark extract experiments, a proposed scheme accurately recognizes the frame more than 97% in total captured images. Also in BER (Bit Error Ratio) of captured images, it shows about 3.73%, then it was improved more than 70%, compared with the Pramila's method.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.27 no.8A
• /
• pp.822-828
• /
• 2002

Optical flow estimation based on multi constraint approaches is frequently used for recognition of moving objects. However, the use have been confined because of OF estimation time as well as error problem. This paper shows a new method form effectively extracting movement information using the multi-constraint base approaches with sobel edge detection. The moving objects anr extraced in the input image sequence using edge detection and segmentation. Edge detection and difference of the two input image sequence gives us the moving objects in the images. The process of thresholding removes the moving objects detected due to noise. After thresholding the real moving objects, we applied the Combinatorial Hough Transform (CHT) and voting accumulation to find the optimal constraint lines for optical flow estimation. The moving objects found in the two consecutive images by using edge detection and segmentation greatly reduces the time for comutation of CHT. The voting based CHT avoids the errors associated with least squares methods. Calculation of a large number of points along the constraint line is also avoided by using the transformed slope-intercept parameter domain. The simulation results show that the proposed method is very effective for extracting optical flow vectors and hence recognizing moving objects in the images.

• Journal of radiological science and technology
• /
• v.42 no.2
• /
• pp.113-117
• /
• 2019

The purpose of this study was to investigate the FA value which can produce the best T2-weighted images by measuring the signal intensity and noise according to the FA value change in the brain image and the abdominal image of the mouse using micro-MRI. Brain imaging and abdominal imaging of BALB / C mice weighing 20g were performed using 4.7T (Bruker BioSpin MRI GmbH) micro-MRI equipment, Turbo RARE-T2 (spin echo-T2) images were scanned at TR 3500 msec and TE 36 msec. The changes of the FA values were $60^{\circ}$, $80^{\circ}$, $100^{\circ}$, $120^{\circ}$, $140^{\circ}$, $160^{\circ}$ and $180^{\circ}$. We measured signal intensity according to FA values of ventricle and thalamus in brain imaging, The signal intensity of kidney and muscle around the kidney was measured in abdominal images. To obtain SNR and CNR, we measured the background signals of two different parts, not the tissue. In the brain (thalamus) image, the signal intensity of FA $100^{\circ}$ was 7,433 and SNR (6.49) was the highest. In the abdominal (kidney) image, the signal intensity was highest at 16,523 when FA was $120^{\circ}$, and the highest SNR was 8.54 when FA was $140^{\circ}$. The CNR value of the brain image was 1.38 at FA $60^{\circ}$ and gradually increased to 8.29 at FA $180^{\circ}$. The CNR value of the muscle adjacent to the kidney gradually increased from 2.36 when the FA value was $60^{\circ}$ and the highest value was 4,57 at the FA value $180^{\circ}$.

• Progress in Medical Physics
• /
• v.15 no.2
• /
• pp.112-119
• /
• 2004

We tried to build human models based on medical images of live Korean, instead of using standard data of human body structures. Characteristics of mechanical structures of human bodies were obtained from medical images such as CT and MR images. For each constitutional part of mechanical structures CT images were analyzed in terms of gray levels and MR images were analyzed in terms of pulse sequence. Characteristic features of various mechanical structures were extracted from the analyses. Based on the characteristics of each structuring element we peformed image segmentation on CT and MR images. We delineated bones, muscles, ligaments and tendons from CT and MR images using image segmentation or manual drawing. For the image segmentation we compared the edge detection method, region growing method and intensity threshold method and applied an optimal compound of these methods for the best segmentation results. Segmented mechanical structures of the head/neck part were three dimensionally reconstructed.